← Back to the dossier

16–17 November 2026 · Isomorphic Labs · OpenAI

5 billion for AI-designed medicines, and 98 percent of scientists expect to discover faster

Isomorphic Labs, the drug discovery company spun out of DeepMind in 2021, raised 5 billion dollars in a Series C on 17 November, three and a half months after Demis Hassabis stepped back as DeepMind's chief executive to run it. It is the company's second round this year, after 2.1 billion in May. The same week OpenAI published a study of 1,819 scientists: 98 percent of those who answered expect AI to speed up discovery in their field, and they estimate the same work would have taken 1.9 times as long without it.

CapabilityConfirmedMore than one independent source, or a primary document.

What happened

The round was co-led by Greenoaks and Sequoia, with Andreessen Horowitz and Alphabet also investing, CNBC reports. Existing backers including GV, CapitalG, Thrive, MGX, Temasek and the UK Sovereign AI Fund joined. TechCrunch puts the two rounds together at 7.1 billion dollars this year. Isomorphic did not disclose its valuation; Bloomberg reported earlier this month that it was seeking about 50 billion, a figure I have not read past the paywall.

What the company says it has is a design engine built on a family of simulation models that explore chemistry, plus programmes in oncology and immunology with what TechCrunch calls positive preclinical results. It has deals with Eli Lilly, Novartis and Johnson & Johnson. No AI-designed drug has regulatory approval, and Isomorphic does not claim one.

The money is not isolated. CNBC reports that drug discovery startups raised 23.2 billion dollars in 2026, a record, citing Dealroom. Anthropic launched a drug discovery programme in June and OpenAI partnered with Novo Nordisk in April. The field has become the clearest commercial test of whether a model can do something that a laboratory then has to confirm.

OpenAI's study is the other half of the same claim, from the bottom up. Its economic research team looked at more than 580,000 scientific messages from 1,819 researchers in a summer pilot and surveyed 251 of them. 68 percent used ChatGPT to develop research ideas; in 35 percent of those messages the scientist brought the question, in 36 percent they refined their own idea, and in 23 percent they asked the model to suggest one. OpenAI also says it is expanding free access to more than 200,000 researchers.

How it workedtechnical detail

Isomorphic's claim is narrower than "AI invents a drug". The engine proposes molecules that fit a biological target, and simulation filters them before anything is made. The expensive part of drug discovery is not the idea; it is the years of synthesis and trials that follow, and most candidates fail there. A model can shrink the first step and cannot touch the last one.

The 1.9 times figure is self-reported by the scientists in the survey, about work they did not do twice. It measures the feeling of speed, not the speed. The honest comparison would be two groups given the same task, one with the model and one without, measured on output. That study does not exist yet, and OpenAI says so: whether the expectation turns into faster discovery will take more research.

What may follow

This is the version of the AI story that does not fit a headline about collapse: money, models and a laboratory that still has to say yes. It is also the version that will be judged first, because a drug either works in a trial or it does not. If one of these programmes reaches a phase 3 readout, the argument about whether models can do science ends differently than the argument about whether they can write code.

What I do not know

No AI-designed medicine has been approved, and the preclinical results are the company's own description. The valuation is undisclosed and the 50 billion figure is a report I could not read. The survey covers people who chose to join a pilot, which is not a random sample of scientists. I did not read the full OpenAI paper, only the summary and the key figures.

Read next

Sources

  1. CNBC: Demis Hassabis' AI company raises $5 billion, months after he left as Google DeepMind CEOpress · main source

    Kai Nicol-Schwarz, 17 November 2026. The round and its lead investors, Hassabis stepping back three and a half months ago, the 2.1 billion Series B in May, the 23.2 billion raised by the sector per Dealroom, and the Anthropic and Novo Nordisk comparisons.

  2. TechCrunch: Isomorphic Labs raises $5B as it looks to take drug candidates to clinical trialspress

    Ram Iyer, 17 November 2026. The 7.1 billion combined, the positive preclinical results in oncology and immunology, the Lilly, Novartis and J&J deals, and the undisclosed valuation with Bloomberg's 50 billion report.

  3. OpenAI: Science at the frontierprimary

    16 November 2026. 580,000 messages from 1,819 scientists, a survey of 251, the 1.9 times estimate, 98 percent expecting faster discovery, the 68 percent ideation figure and its split, and the expansion to 200,000 researchers.

Comments